Trade Matrix Labs

Athena AI Finance turns a dense SEC financial report into a AI-agent native customized report a reader can check. Filings are long and hard to read: institutions run short of time, individuals often lack the training or the language, and a general AI can change its numbers or make them up. In Athena, a reader opens a stored report and clicks any linked figure to see whether it is quoted or calculated and to open the filing line it came from. Ask, a LangChain agent on Muse Spark 1.3, answers from that report only, cites its passages, checks the figures it writes against them, and waits for the reader's approval before any change. Memos and chats stay with the reader in the Library. New in this hackathon: the reading app(Athena AI Web-App), the local server and database with code checks at ingest, Ask, the number checker and the Ask evaluation. Earlier: the idea, the target report quality, a private draft report generator in research and testing, and UI mockups. The sample report is a draft held as review required. Code checks prove references, not meaning. Not investment advice.